Proud of my clever agents--and a need for better bequests

1

A few months ago I lamented the behavior of my agents (Deadbeat agents and feasibility constraints). I’ve gotten better at not putting them into impossible positions, and they’ve gotten better at surprising me with their optimal solution-finding. The latest shows their decision-making when they have an option to purchase a solar endowment in the face of ever-rising fossil-fuel prices:

SolarEndowment
SolarEndowment1741×1126 37.2 KB

Previously, they would accumulate assets, putting them into high-yield shares (which are quantized like bonds, so they put small-money into low-yield savings), then when they had enough money, they’d buy a house for the benefits of receiving better housing services, then they’d sell the house and live off the proceeds until their died. Everything was hump-shaped.

Now, in the face of relentless increases in fossil fuels over time, they choose to buy their house, install solar, and then benefit from the endowment of self-generated power that greatly reduces their exposure to increasing energy costs. Clever agents!

Now I have a new problem: I currently follow a simplistic model whereby shares, assets, and houses feed into accidental bequests at their market value, with beneficiaries receiving the (small) average of those bequests. Those bequests satisfy the consumption formula, which allows agents to otherwise use accumulated assets for other purposes, such as house and solar PV acquisition. But in reality, some agent could inherit the house and its solar endowment, and this is not a “decision” per se. But perhaps this could be modeled as a kind of a shock which, when observed by an agent, makes the “decision” of buying a fully equipped house with some amount of solar PV capacity a no-cost option cost.

Presently I used the buyhouse decision function from LifeCycleModel35. I could add a purchase type of “inherit” which would only be satisfied if the shock type and value indicated they were the beneficiary. Is that a good way to think about it? Is it worth considering the case whereby real property is bequested, or do most economists argue (existentially?) that one will get essentially the same answer by converting everything to cash and distributing that?

1 Like
2

I expect most Economists would argue that for the individual, there is little to no difference unless you think transaction costs are huge (what is the difference between $1million and a house worth $1million, it is not a big effort to convert one into the other).

Of course, at the level of the whole economy this is not true. You cannot just evaporate houses into piles of cash.

From the model perpective, you can count the ‘dollar bequests left’ and the ‘house bequests left’ separately, and then people can receive these, and you make them each add up in general eqm (one ‘bequests left equal bequests received for each of dollars and houses’). The most obvious challenge will be that people who already have a house cannot receive another one (guessing about model state space), and you don’t want people not buying houses because they want to avoid being unable to inherit one if they already bought one.

Easiest is probably just do the ‘all inheritance is dollars’ approach and then just have the ‘no evaporation of houses’ enforced by a general eqm in housing market (not sure if your model imposes housing supply in some way).

3

I do use CustomModelStats to discover housing supply using LifeCycleAnalysis. As usual, all agents start with no assets anywhere on the grid. As they accumulate assets (with age), that changes. At each candidate GE solution, and across transition points, I can count the home purchases and count the home sales, and from there, net out the actual needed new supply (rental houses are invisible and in infinite supply).

Houses cost between 4x and 12x salary, which does cover the example of a $1M+ home. You are correct that we don’t allow the ownership of multiple houses, but we do allow differently-sized houses, so one could inherit a large house and on-sell a smaller one that is owned.

I need to check whether houses held at death are treated as a sale (for estate purposes) or an evaporation-to-cash. But I’m not so intent on modeling housing stock per se, just household, firm, and energy company behavior.